A Mult-Task System for Detecting and Classifying Malware Signatures Using Random Forest Classifier

Abstract

The rapid increase in the use of information technology has made cyber-attacks a major concern in the use of internet by users globally. These attacks are carried out in different forms, some are carried out as phishing, man in the middle, malicious applications and so on. In this study we will focus on malware attack. Malicious applications have been a major challenge in the use of applications on windows operating system. These malicious attacks are being carried out in different forms. Some of these attacks are trojan, ransom, keylogger etc. The need to detect and classifier these malicious attacks in windows operating system is an important task. So therefore, this paper presents a smart system for detecting and classifying eight categories of malware attack on windows operating system using random forest classifier. The system starts by collecting signatures of malware attack on windows from Virus Share, Virus Sign and Github respiratory. The collected malware signatures went through the following stages of preprocessing (First stage, Second Stage, and Third Stage). The first stage has to do with creating a pandas. Dataframe using the malware signatures. The second stage has to with data cleaning and the third stage has to do with data transformation. The result of the Random Forest Classifier shows a promising performance in terms of accuracy, precision, f1-score, and recall. The result shows that the Random Forest Classifier has an accuracy of about 100% for each of the matrix evaluation. Keywords- Malware signatures, Random Forest Classifier, Windows operating System, Matrix Evaluation